Real Time Conditioning Monitoring for Failure Prediction
Nowadays, the growth in the diameter of the rotors and the power capacity of the machinery require the application of constant monitoring to predict failures and reduce maintenance activities. These activities prevent emergency stops and failures that require the replacement of components; however, these operations involve a high cost. The system presented in this paper integrates a fast data processing method that can analyze monitoring signals in real-time. The overall system includes a set of sensors for the measurement of acceleration, angular velocity, temperature, voltage, electrical current, and oil quality. The system presented is able to produce a frequency spectrum with a length of 8192 points. The FFT module was implemented in a FPGA (Field Programmable Gate Array) and it generates in real time, the frequency spectrum. This system adjusts the dominant frequencies to the wind turbine velocity.